As the depth of China’s reliance on its old stand by of investment growth fueled by increasingly risky credit becomes more apparent, Beijing and China bulls have fallen back on the old standby citing Chinese rebalancing. Given the seeming rapid growth in metrics like investment and credit, it becomes important to unpack whether China actually is moving away from its historical growth model.
There is top line data supporting the idea that China is rebalancing away from its investment heavy model. For instance in Q1 2010, secondary industry comprise 56.4% of the Chinese economy but has fallen steadily since then to 37.2% or by nearly 20% of GDP while the tertiary sector has seen a nearly identical corresponding increase. Of the 6.7% GDP increase 3.9% of that supposedly came from the tertiary sector.
If we look at other related numbers, there are other top line numbers which support this idea that China is rebalancing. According to official data, even now retail sales only recently dropped beneath to hit 9.5% or nearly 3% more than real GDP growth. This sounds like a nearly air tight case for rebalancing right? As I always say, move past the headline data and you get a very different picture of what is actually happening in the Chinese economy.
Let’s start with the reliance on investment and credit to drive growth assuming for as long as possible the data is accurate. In 2009, fixed asset investment (FAI) was equal to 56% of nominal GDP while in 2016 it was equal to 80% of nominal GDP and 80% in 2015. However, from gross capital formation (GCF), the GDP accounting representation with some important exclusions, of FAI shows a different patter. In 2009, GCF was equal to 46% of nominal GDP while as FAI was rising rapidly through 2015, GCF actually dropped as a percentage of GDP to 45%. In other words, while the cash value of investment in the Chinese economy has been rising rapidly since 2009, its GDP measure has actually dropped.
We are now left with a conundrum: is it possible to reconcile the rapid growth in FAI with the drop in GCF within Chinese GDP statistics? Possible but extremely unlikely and even if so leaves Chinese finances in an vastly more precarious position. The primary exclusion between FAI, the financial cost of investment, and GCF, the GDP accounting measure of investment, is the value of land is excluded. In 2015, FAI was 80% of GDP while GCF was only 45% so this raises the question whether land sales in investment comprised 35% of GDP and whether the value of land sales have risen dramatically this decade?
Looking at the data it is very difficult to see how land value contributes to this supposed wedge or any major increase in the sales of land values. In 2010, total land sales values in 100 large and medium sized cities was 1.8 trillion RMB. As a point of comparison, in 2010 total FAI was 24.1 trillion RMB or 8% of the total. Given that GCF in 2010 is counted as 19.7 trillion RMB, it is not inconceivable that these numbers reconcile close enough for our purposes. The difference between FAI, 100 city land value, and GCF (FAI-100 City Land – GCF) is only 2.6 trillion RMB. Given cities and areas outside major urban centers, it is not inconceivable that we could approach that 2.6 trillion RMB level.
However, since then this line of reasoning in the numbers fall apart. Between 2010 and 2015, the last year we have GCF data for, total GCF has risen 59% and FAI 128%. That implies that the wedge between GCF and FAI is due to rapidly rising land value sales. As just noted, in 2010 total land sales value in 100 large and medium cities was 1.84 trillion RMB; in 2015, total land sales values in 100 large and medium cities was 1.81 trillion RMB. The wedge between FAI and GCF in 2015 is now a much more substantial 23.9 trillion RMB. If we subtract out the value of land sold in 100 large and medium cities of 1.8 trillion RMB, this still leaves a wedge of 22.03 trillion.
This raises a number of key points. First, it stretches credibility well beyond the breaking point to believe that rural and small cities sold land equal to 32% of nominal GDP in 2015. One way we can see this is that real estate FAI simply is not that large. The entirety of real estate FAI in in 2015 was 9.6 trillion but somehow magically land sales in China are supposed to represent nearly 24 trillion that year. It is a mathematical impossibility that both of those numbers are true.
Second, if the land sales wedge that might explain the FAI and GCF wedge is effectively non-existent, this implies that gross capital formation is radically undervalued in GDP statistics. Take a simple assumption that rather than dropping as a share of GDP, that GCF grew more in line with FAI. For our purposes, assume that rather than the 45% it now represents or the 81% share of GDP that FAI represents, assume we split the difference. That means that GCF as a share of Chinese GDP would now represent a staggering 63% of GDP. Today by official numbers GCF represents 37% of GDP and has never topped 60% since 2010. To put this number in perspective, according to the World Bank, only 9 countries had GCF as a percentage of GDP about 40% and only economic powerhouse Suriname was above 60%. In other words, whether you choose to believe the official GCF data or believe that GCF is somewhere closer to FAI, whatever that exact number, China remains as grossly unbalanced country, the only question is how unbalanced exactly.
Third, the next question is whether FAI data is potentially overstated. If we compare FAI data to various forms of financing like total social financing to the real economy, we see a big discrepancy. FAI is significantly higher than TSF and has grown much faster over time. In 2010, FAI in China amounted to 25 trillion RMB while TSF was 14 trillion. By 2016, those numbers had become 60.6 trillion and 17.8 trillion. In other words, somehow FAI increased by 35 trillion while TSF increased by only 3.8 trillion or by a tenth of FAI. That may seem like an open and shut case that FAI is overstated. However, it isn’t.
Many industries who provide the inputs for FAI activity revenue grows much more in line with the FAI growth than with financing. Nonmetallic mineral manufacturing (think cement, glass, etc) nearly doubled their revenue growth from 2010 to 2015 while FAI was a little above that at 120% even as yearly TSF only grew 10% during that same time. Other industries like specialty purpose machinery and nonferrous metals grew by very similar amounts indicating there is a much closer relationship to FAI than financing metrics like TSF.
This then has two further implications. First, it seems to imply that there might be hidden financing pushing FAI as it is not statistically at least coming from TSF. Second, it implies that FAI is a much more accurate portrayal of the Chinese economy’s reliance on investment for growth than GCF. By virtually any real adjustment, this means the Chinese economy is more unbalanced than almost any other time in modern history.
There is one final point here. Assume for one minute that the wedge between FAI and GCF is entirely explainable by land sales in China which is subsequently being used to finance this gap. A tenuous assumption but work with me. This means Chinese public finances are increasingly fragile on many many levels. For instance, for the government to raise taxes to a level to entirely or large replace land sales, which verifiably account in many places for 50% of government revenue, they would need to raise taxes to astounding levels. It further raises the specter that Chinese governments have to increase land sales at every increasing rates. Furthermore, it implies that there is so enormous level of “hidden” debt that simply isn’t being accounted for to fund land sales on this scale to the tune of roughly $3 trillion USD yearly. If this is true, this is truly terrifying for financial stability.
No matter how you look at it, looking at investment and credit, the Chinese economy is more reliant on investment and credit to fund growth than ever before.